Utilities

Output Parser

Replace the legacy Output Parser component with the Structured Output component and Parser component. The components you need depend on the data types and complexity of the parsing task.

The Output Parser component transforms the output of a language model into comma-separated values (CSV) format, such as [“item1”, “item2”, “item3”], using CommaSeparatedListOutputParser. The Structured Output component is a good alternative for this component because it also formats LLM responses with support for custom schemas and more complex parsing.

Parsing components only provide formatting instructions and parsing functionality. They don’t include prompts. You must connect parsers to Prompt Template components to create prompts that LLMs can use.

1. Open a flow that has a Chat Input, Language Model, and Chat Output components. 
2. Add Output Parser and Prompt Template components to your flow. 
3. Define your LLM’s prompt in the Prompt Template component’s Template, including all instructions and pre-loaded context. Make sure to include a {format_instructions} variable where you will inject the formatting instructions from the Output Parser component. For example:

You are a helpful assistant that provides lists of information.

Variables in the template dynamically add fields to the Prompt Template component so that your flow can receive definitions for those values from other components, Robility flow global variables, or fixed input.

4. Connect the Output Parser component’s output to the Prompt Template Component’s format instructions input.

The Output Parser component accepts the following parameters: